install DuckDB when running benchmarks (#541)

Fix failures on benchmark action on develop

.github/workflows/bench.yml

@@ -12,7 +12,6 @@ permissions:
 jobs:
   bench:
     runs-on: ubuntu-latest-large
-    if: ${{ github.ref_name == 'develop' }}
     steps:
       - uses: actions/checkout@v4
         with:
@@ -28,6 +27,12 @@ jobs:
       - name: Setup bencher.dev CLI
         uses: bencherdev/bencher@main
 
+      # The compression benchmarks rely on DuckDB being installed to convert CSV to Parquet
+      - name: Install DuckDB
+        uses: opt-nc/[email protected]
+        with:
+          version: v1.0.0
+
       - name: Upload Vortex Benchmarks
         run: |
           bencher run \
@@ -38,4 +43,10 @@ jobs:
             "cargo bench -p bench-vortex"
         env:
           BENCHER_TOKEN: ${{ secrets.BENCHER_TOKEN }}
+          # AWS Credentials for R2 storage tests
+          AWS_BUCKET: vortex-test
+          AWS_ENDPOINT: ${{ secrets.AWS_ENDPOINT }}
+          AWS_ACCESS_KEY_ID: ${{ secrets.AWS_ACCESS_KEY_ID }}
+          AWS_SECRET_ACCESS_KEY: ${{ secrets.AWS_SECRET_ACCESS_KEY }}
+
 

Cargo.toml

@@ -75,6 +75,7 @@ futures-util = "0.3.30"
 getrandom = "0.2.14"
 half = { version = "^2", features = ["std", "num-traits"] }
 hashbrown = "0.14.3"
+homedir = "0.3.3"
 humansize = "2.1.3"
 indicatif = "0.17.8"
 itertools = "0.13.0"
@@ -104,6 +105,7 @@ serde = "1.0.197"
 serde_json = "1.0.116"
 serde_test = "1.0.176"
 simplelog = { version = "0.12.2", features = ["paris"] }
+tar = "0.4"
 thiserror = "1.0.58"
 tokio = "1.37.0"
 uninit = "0.6.2"

bench-vortex/Cargo.toml

@@ -26,6 +26,7 @@ datafusion = { workspace = true }
 enum-iterator = { workspace = true }
 flexbuffers = { workspace = true }
 futures = { workspace = true, features = ["executor"] }
+homedir = { workspace = true }
 humansize = { workspace = true }
 indicatif = { workspace = true }
 itertools = { workspace = true }
@@ -40,6 +41,7 @@ rayon = { workspace = true }
 reqwest = { workspace = true }
 serde = { workspace = true }
 simplelog = { workspace = true }
+tar = { workspace = true }
 tokio = { workspace = true, features = ["full"] }
 uuid = { workspace = true, features = ["v4"] }
 vortex-alp = { path = "../encodings/alp" }